Fetching the paper…

Using Human Feedback to Fine-tune Diffusion Models without Any Reward Model · Around